Glycotherapeutics
Protein-based therapeutics currently represent one in every four new drugs approved by the FDA (Walsh, G., “Biopharmaceutical Benchmarks,” Nat Biotechnol 18:831-3 (2000); Walsh, G, “Biopharmaceutical Benchmarks,” Nat Biotechnol 21:865-70 (2003); and Walsh, G, “Biopharmaceutical Benchmarks,” Nat Biotechnol 24:769-76 (2006)).
While several protein therapeutics can be produced using a prokaryotic expression system such as E. coli (e.g., insulin), the vast majority of therapeutic proteins require additional post-translational modifications, thought to be absent in prokaryotes, to attain their full biological function. In particular, N-linked protein glycosylation is predicted to affect more than half of all eukaryotic protein species (Apweiler et al., “On the Frequency of Protein Glycosylation, as Deduced From Analysis of the SWISS-PROT Database,” Biochim Biophys Acta 1473:4-8 (1999)) and is often essential for proper folding, pharmacokinetic stability, tissue targeting and efficacy for a large number of proteins (Helenius et al., “Intracellular Functions of N-linked Glycans,” Science 291:2364-9 (2001)). Since most bacteria do not glycosylate their own proteins, expression of most therapeutically relevant glycoproteins, including antibodies, is relegated to mammalian cells. However, mammalian cell culture suffers from a number of drawbacks including: (i) extremely high manufacturing costs and low volumetric productivity of eukaryotic hosts, such as CHO cells, relative to bacteria; (ii) retroviral contamination; (iii) the relatively long time required to generate stable cell lines; (iv) relative inability to rapidly generate stable, “high-producing” eukaryotic cell lines via genetic modification; and (v) high product variability created by glycoform heterogeneity that arises when using host cells, such as CHO, that have endogenous non-human glycosylation pathways (Choi et al., “Use of Combinatorial Genetic Libraries to Humanize N-linked Glycosylation in the Yeast Pichia pastoris,” Proc Natl Acad Sci USA 100:5022-7 (2003)). Expression in E. coli, on the other hand, does not suffer from these limitations.
Expression of a Glycosylated Therapeutic Proteins in E. coli 
Many therapeutic recombinant proteins are currently expressed using E. coli as a host organism. One of the best examples is human insulin, which was first produced in E. coli by Eli Lilly in 1982. Since that time, a vast number of human therapeutic proteins have been approved in the U.S. and Europe that rely on E. coli expression, including human growth hormone (hGH), granulocyte macrophage colony stimulating factor (GM-CSF), insulin-like growth factor (IGF-1, IGFBP-3), keratinocyte growth factor, interferons (IFN-α, IFN-β1b, IFN-γ1b), interleukins (IL-1, IL-2, IL-11), tissue necrosis factor (TNF-α), and tissue plasminogen activator (tPA). However, almost all glycoproteins are produced in mammalian cells. When a protein that is normally glycosylated is expressed in E. coli, the lack of glycosylation in that host can yield proteins with impaired function. For instance, aglycosylated human monoclonal antibodies (mAbs) (e.g., anti-tissue factor IgG1) can be expressed in soluble form and at high levels in E. coli (Simmons et al., “Expression of Full-length Immunoglobulins in Escherichia coli: Rapid and Efficient Production of Aglycosylated Antibodies,” J Immunol Methods 263:133-47 (2002)). However, while E. coli-derived mAbs retained tight binding to their cognate antigen and neonatal receptor and exhibited a circulating half-life comparable to mammalian cell-derived antibodies, they were incapable of binding to C1q and the FcγRI receptor due to the absence of N-glycan.
Eukaryotic and Prokaryotic N-Linked Protein Glycosylation
N-linked protein glycosylation is an essential and conserved process occurring in the endoplasmic reticulum (ER) of eukaryotic organisms (Burda et al., “The Dolichol Pathway of N-linked Glycosylation,” Biochim Biophys Acta 1426:239-57 (1999)). It is important for protein folding, oligomerization, quality control, sorting, and transport of secretory and membrane proteins (Helenius et al., “Intracellular Functions of N-linked Glycans,” Science 291:2364-9 (2001)). The eukaryotic N-linked protein glycosylation pathway (FIG. 1) can be divided into two different processes: (i) the assembly of the lipid-linked oligosaccharide at the membrane of the endoplasmic reticulum and (ii) the transfer of the oligosaccharide from the lipid anchor dolichyl pyrophosphate to selected asparagine residues of nascent polypeptides. The characteristics of N-linked protein glycosylation, namely (i) the use of dolichyl pyrophosphate (Dol-PP) as carrier for oligosaccharide assembly, (ii) the transfer of only the completely assembled Glc3Man9GlcNAc2 oligosaccharide, and (iii) the recognition of asparagine residues characterized by the sequence N-X-S/T where N is asparagine, X is any amino acid except proline, and S/T is serine/threonine (Gavel et al., “Sequence Differences Between Glycosylated and Non-glycosylated Asn-X-Thr/Ser Acceptor Sites: Implications for Protein Engineering,” Protein Eng 3:433-42 (1990)) are highly conserved in eukaryotes. The oligosaccharyltransferase (OST) catalyzes the transfer of the oligosaccharide from the lipid donor dolichylpyrophosphate to the acceptor protein. In yeast, eight different membrane proteins have been identified that constitute the complex in vivo (Kelleher et al., “An Evolving View of the Eukaryotic Oligosaccharyltransferase,” Glycobiology 16:47R-62R (2006)). STT3 is thought to represent the catalytic subunit of the OST (Nilsson et al., “Photocross-linking of Nascent Chains to the STT3 Subunit of the Oligosaccharyltransferase Complex,” J Cell Biol 161:715-25 (2003) and Yan et al., “Studies on the Function of Oligosaccharyl Transferase Subunits. Stt3p is Directly Involved in the Glycosylation Process,” J Biol Chem 277:47692-700 (2002)). It is the most conserved subunit in the OST complex (Burda et al., “The Dolichol Pathway of N-linked Glycosylation,” Biochim Biophys Acta 1426:239-57 (1999)).
Conversely, the lack of glycosylation pathways in bacteria has greatly restricted the utility of prokaryotic expression hosts for making therapeutic proteins, especially since by certain estimates “more than half of all proteins in nature will eventually be found to be glycoproteins” (Apweiler et al., “On the Frequency of Protein Glycosylation, as Deduced From Analysis of the SWISS-PROT Database,” Biochim Biophys Acta 1473:4-8 (1999)). Recently, however, it was discovered that the genome of a pathogenic bacterium, C. jejuni, encodes a pathway for N-linked protein glycosylation (Szymanski et al., “Protein Glycosylation in Bacterial Mucosal Pathogens,” Nat Rev Microbiol 3:225-37 (2005)). The genes for this pathway, first identified in 1999 by Szymanski and coworkers (Szymanski et al., “Evidence for a System of General Protein Glycosylation in Campylobacter jejuni,” Mol Microbiol 32:1022-30 (1999)), comprise a 17-kb locus named pgl for protein glycosylation. Following discovery of the pgl locus, in 2002 Linton et al. identified two C. jejuni glycoproteins, PEB3 and CgpA, and showed that C. jejuni-derived glycoproteins such as these bind to the N-acetyl galactosamine (GalNAc)-specific lectin soybean agglutinin (SBA) (Linton et al., “Identification of N-acetylgalactosamine-containing Glycoproteins PEB3 and CgpA in Campylobacter jejuni,” Mol Microbiol 43:497-508 (2002)). Shortly thereafter, Young et al. identified more than 30 potential C. jejuni glycoproteins, including PEB3 and CgbA, and used mass spectrometry and NMR to reveal that the N-linked glycan was a heptasaccharide with the structure GalNAc-α1,4-GalNAc-α1,4-[Glcβ1,3]GalNAc-α1,4-GalNAc-α1,4-GalNAc-α1,3-Bac-β1,N-Asn (GalNAc5GlcBac, where Bac is bacillosamine or 2,4-diacetamido-2,4,6-trideoxyglucose) (Young et al., “Structure of the N-linked Glycan Present on Multiple Glycoproteins in the Gram-negative Bacterium, Campylobacter jejuni,” J Biol Chem 277:42530-9 (2002)) (FIG. 2). The branched heptasaccharide is synthesized by sequential addition of nucleotide-activated sugars on a lipid carrier undecaprenylpyrophosphate on the cytoplasmic side of the inner membrane (Feldman et al., “Engineering N-linked Protein Glycosylation with Diverse O Antigen Lipopolysaccharide Structures in Escherichia coli,” Proc Natl Acad Sci USA 102:3016-21 (2005)) and, once assembled, is flipped across the membrane by the putative ATP-binding cassette (ABC) transporter WlaB (Alaimo et al., “Two Distinct But Interchangeable Mechanisms for Flipping of Lipid-linked Oligosaccharides,” Embo J 25:967-76 (2006) and Kelly et al., “Biosynthesis of the N-linked Glycan in Campylobacter jejuni and Addition Onto Protein Through Block Transfer,” J Bacteriol 188:2427-34 (2006)). Next, transfer of the heptasaccharide to substrate proteins in the periplasm is catalyzed by an OST named PglB, a single, integral membrane protein with significant sequence similarity to the catalytic subunit of the eukaryotic OST STT3 (Young et al., “Structure of the N-linked Glycan Present on Multiple Glycoproteins in the Gram-negative Bacterium, Campylobacter jejuni,” J Biol Chem 277:42530-9 (2002)). PglB attaches the heptasaccharide to asparagine in the motif D/E-X1-N-X2-S/T (where D/E is aspartic acid/glutamic acid, X1 and X2 are any amino acids except proline, N is asparagine, and S/T is serine/threonine), a sequon similar to that used in the eukaryotic glycosylation process (N-X-S/T) (Kowarik et al., “Definition of the Bacterial N-glycosylation Site Consensus Sequence,” Embo J 25:1957-66 (2006)).
Glycoengineering of Microorganisms
A major problem encountered when expressing therapeutic glycoproteins in mammalian, yeast, or even bacterial host cells is the addition of non-human glycans. For instance, yeast, one of the two most frequently used systems for the production of therapeutic glycoproteins, transfer highly immunogenic mannan-type N-glycans (containing up to one hundred mannose residues) to recombinant glycoproteins. Mammalian expression systems can also modify therapeutic proteins with non-human sugar residues, such as the N-glycosylneuraminic acid (Neu5Gc) form of sialic acid (produced in CHO cells and in milk) or the terminal α(1,3)-galactose (Gal) (produced in murine cells). Repeated administration of therapeutic proteins carrying non-human sugars can elicit adverse reactions, including an immune response in humans.
As an alternative to using native glycosylation systems for producing therapeutic glycoproteins, the availability of glyco-engineered expression systems could open the door to customizing the glycosylation of a therapeutic protein and could lead to the development of improved therapeutic glycoproteins. Such a system would have the potential to eliminate undesirable glycans and perform human glycosylation to a high degree of homogeneity. To date, only the yeast Pichia pastoris has been glyco-engineered to provide an expression system with the capacity to control and optimize glycosylation for specific therapeutic functions (Gerngross, T. U., “Advances in the Production of Human Therapeutic Proteins in Yeasts and Filamentous fungi,” Nat Biotechnol 22:1409-14 (2004); Hamilton et al., “Glycosylation Engineering in Yeast: The Advent of Fully Humanized Yeast,” Curr Opin Biotechnol 18:387-92 (2007); and Wildt et al., “The Humanization of N-glycosylation Pathways in Yeast,” Nat Rev Microbiol 3:119-28 (2005)).
For example, a panel of glyco-engineered P. pastoris strains was used to produce various glycoforms of the monoclonal antibody Rituxan (an anti-CD20 IgG1 antibody) (Li et al., “Optimization of Humanized IgGs in Glycoengineered Pichia pastoris,” Nat Biotechnol 24:210-5 (2006)). Although these antibodies share identical amino acid sequences to commercial Rituxan, specific glycoforms displayed ˜100-fold higher binding affinity to relevant FcγRIII receptors and exhibited improved in vitro human B-cell depletion (Li et al., “Optimization of Humanized IgGs in Glycoengineered Pichia pastoris,” Nat Biotechnol 24:210-5 (2006)). The tremendous success and potential of glyco-engineered P. pastoris is not without some drawbacks. For instance, in yeast and all other eukaryotes N-linked glycosylation is essential for viability (Herscovics et al., “Glycoprotein Biosynthesis in Yeast,” FASEB J 7:540-50 (1993) and Zufferey et al., “STT3, a Highly Conserved Protein Required for Yeast Oligosaccharyl Transferase Activity In Vivo,” EMBO J 14:4949-60 (1995)). Thus, the systematic elimination and re-engineering by Gerngross and coworkers of many of the unwanted yeast N-glycosylation reactions (Choi et al., “Use of Combinatorial Genetic Libraries to Humanize N-linked Glycosylation in the Yeast Pichia pastoris,” Proc Natl Acad Sci USA 100:5022-7 (2003)) has resulted in strains that are “sick” compared to their wild-type progenitor. This can be worsened during high-level glycoprotein expression due to the large metabolic burden placed on the yeast glycosylation system. As a result, the cell yield that can be obtained during large-scale fermentation is limited. Furthermore, elimination of the mannan-type N-glycans is only half of the glycosylation story in yeast. This is because yeast also perform O-linked glycosylation whereby O-glycans are linked to Ser or Thr residues in glycoproteins (Gentzsch et al., “The PMT Gene Family: Protein O-glycosylation in Saccharomyces cerevisiae is Vital,” EMBO J 15:5752-9 (1996)). As with N-linked glycosylation, O-glycosylation is essential for viability (Gentzsch et al., “The PMT Gene Family: Protein O-glycosylation in Saccharomyces cerevisiae is Vital,” EMBO J 15:5752-9 (1996)) and thus cannot be genetically deleted from glyco-engineered yeast. Since there are differences between the O-glycosylation machinery of yeast and humans, the possible addition of O-glycans by glyco-engineered yeast strains has the potential to provoke adverse reactions including an immune response.
Recently, Aebi and his coworkers transferred the C. jejuni glycosylation locus into E. coli and conferred upon these cells the extraordinary ability to post-translationally modify proteins with N-glycans (Wacker et al., “N-linked Glycosylation in Campylobacter jejuni and its Functional Transfer into E. coli,” Science 298:1790-3 (2002)). However, despite the functional similarity shared by the prokaryotic and eukaryotic glycosylation mechanisms, the oligosaccharide chain attached by the prokaryotic glycosylation machinery (GalNAc5GlcBac) is structurally distinct from that attached by eukaryotic glycosylation pathways (Szymanski et al., “Protein Glycosylation in Bacterial Mucosal Pathogens,” Nat Rev Microbiol 3:225-37 (2005); Young et al., “Structure of the N-linked Glycan Present on Multiple Glycoproteins in the Gram-negative Bacterium, Campylobacter jejuni,” J Biol Chem 277:42530-9 (2002); and Weerapana et al., “Asparagine-linked Protein Glycosylation: From Eukaryotic to Prokaryotic Systems,” Glycobiology 16:91R-101R (2006)). Numerous attempts (without success) have been made to reprogram E. coli with a eukaryotic N-glycosylation pathway to express N-linked glycoproteins with structurally homogeneous human-like glycans.
The present invention is directed to overcoming the deficiencies in the art.